Study On Measurement Of The Parameters Of Thin Films By Transmission Spectra Method

It is always be studied by thin film workers that how to determinate the thin film parameters with high degree accuracy. Many methods have been proposed in the past to measure the parameters of thin films, and the transmission spectra method is one of the most favorite methods for many people. This is due to many merits such as simplicity of measurement, facility of operation, variety of thin film, high accuracy. But the problem of this method is that is very difficult to solve the parameters directly by use of the experiment data and the multiplicity of solutions.In this paper, I studied transmission spectra method by which parameters of thin films can be determined explicitly and systematically. The transmittance formula included most of the film and substrate parameters, which affected the transmission spectra, was given in a simplified form, and it was rigorously derived from electromagnetic theory. A new method of simultaneous determination the multi-parameters by use of transmission spectra and simulated annealing algorithm has been presented. Thin film parameters were obtained by solving the equations in inversion method from experimental transmission data. On the one hand, it can solve the problem of multiplicity of solutions efficiently; On the other hand, it applies to all kinds of transmission spectra and doesn't rely on the existence of fringe patterns or transparency. Application of this method and discussion were described in the measurement of transparent films and absorbing films. Simulated results and experimental results were satisfied. A set of calculation softwares was programmed by simulated annealing algorithm. These softwares can find the best solution withoutbeing given initial values but region of convergence, and have a lot of advantages such as fast convergence, good stability and high accuracy. This provided a credible calculation tool for us to solve the parameters of thin films that we nearly know anything about, and avoid the error or undependable consequences because unsuitable initial values. At last, effects of random errors in transmission data on the accuracy of determination of the parameters of thin films were analyzed. Computational experiments with simulated random errors showed that these errors were not critical for the accuracy of thin film parameters.